C.V. Malcolm, T.C. Swaan, H.I. Ridings
Dept. of Agriculture of Western Australia, Perth
In many countries there are extensive areas of saline land which can neither be irrigated nor used for growing normal crops and pastures. In a long term Western Australian research programme, techniques for producing grazing from salt-tolerant forage plants sown on saline soils under natural rainfall have been developed (Malcolm, 1974). The possible application of this approach to other countries has been discussed elsewhere (Malcolm, 1969).
The main salt-affected soils are in the cereal and sheep growing areas where the annual rainfall of 300-375 mm is received largely in the cool period of the year from May to September. The summers are hot and dry. The elevation of the saline areas is less than 400 m above sea level and they do not suffer from excessive frosts. Studies (Malcolm, 1979, Malcolm and Clarke, 1971) indicate that Atriplex rhagodioides, A. undulata and Maireana brevifolia are well suited to forage production on saline soils in south-western Australia. The studies were conducted using glasshouse-raised bushes planted into field situations. Labour costs preclude the use of glasshouse-raised bushes for planting broad scale areas in south-western Australia. Therefore a reliable field method of establishment using seed is required.
In environments such as saline soils, waste dumps and arid areas, the establishment of shrubs from seed presents formidable difficulties. The problems of establishing shrubs in saline soils from seed under natural rainfall are reviewed elsewhere (Malcolm, 1971).
The Mallen niche seeder has been developed (Malcolm, in preparation) to overcome these problems. In the present study niche seeding is compared with the previously-used method of spreading seed on cultivated land.
The Mallen niche seeder consists of a cultivation implement, which, in the models used in the experiment, is a mouldboard plough, and an attached planting section. The planting section rides on a press wheel which forms a niche on the bank produced by the plough. The turning of the press wheel drives a mechanism which deposits seed with a covering of chaff in spot placements in the niche. The interval between placements can be adjusted. For experimental purposes a specific number of placements may be sown in a plot and the machine may be calibrated to determine the average output of seed per placement. Results may therefore be expressed either as a percentage of successful spot placements or if all seedlings are counted, as the percentage of viable seeds giving seedlings.
Germination of M. brevifolia occurs readily in petri dishes but the bracts of the two Atriplex spp. inhibit germination. The bracts must therefore be removed for seedfill and germinability to be determined. Seedfill and germination tests were conducted on subsamples of the seed used in the field experiment. The results were used to convert the field calibration figures for the Mallen Niche Seeder to the percentage of viable seeds giving established plants.
During 1976 two detailed experiments were planted with the Mallen niche seeder on a duplex saline soil with saline groundwater ranging from about 0.5 to 1.5 m depth and salinity of the order of 23 000 to 30 000 mg/l total dissolved solids. In the experimental area two sub-sites were selected, a lower-lying more waterlogged and saline site and a less severe area. Both areas had been excluded from cropping owing to salt encroachment and carried a patchy cover dominated by Hordeum marinum and Eragrostis dielsii. There were larger areas bare in the severe than the moderate site.
The treatments used in Experiment A were as follows:
Site: Waterlogged
Non-waterlogged
Time of sowing: Before opening rain (late April)
After opening rain (late May)
Species: M. brevifolia
A. undulata
A. rhagodioides
Sowing method: Mallen seeder with chaff, DDT, PVA
and Rogor Mallen seeder with chaff;
Mallen seeder without chaff;
Seed spread on cultivated ground.
The treatments were combined factorially and replicated four times.
An early sowing is possible on many saline soils because they remain sufficiently moist and soft to be cultivated readily. Early sowing, if successful, has the advantage of not coinciding with the busy seeding period after the opening rain. Later sowing provides an opportunity for some leaching of surface salt to occur before the seeds are sown.
A feature of the Mallen niche seeding technique is that it enables an expensive treatment such as chaff mulch to be applied as a spot treatment, thereby greatly reducing the cost. In the present experiment the chaff in one treatment was treated with DDT and polyvinyl acetate (PVA) emulsion for the respective purposes of reducing insect predation and helping the chaff remain in position in the niche. The PVA treated chaff tends to stick together once applied. Rogor was applied to the seed in this treatment in a further attempt to reduce adverse effects of insects. A scarifier was used to cultivate the conventional plots; a weighed amount of seed being spread by hand over a standard area of the plots.
In the second experiment, B, using the same three species, the following treatments were used either with or without Rogor treatment of the seed.
Mallen niche seeder without chaff
Mallen niche seeder with chaff
Mallen niche seeder with chaff + PVA
Mallen niche seeder with chaff + DDT
Mallen niche seeder with chaff + DDT + PVA
The second experiment was planted after the opening rain in the less severe section of the experimental area. The treatments and species were combined factorially and replicated three times.
The results of the two experiments are summarised in Tables 1 and 2. A feature of the results in Experiment A is the variability. The coefficient of variation ranged from 77.5% on the non-waterlogged site to 90.2% for the whole experiment. Nevertheless significant treatment differences were demonstrated.
Table 1 Mean percentages of establishment of viable seeds planted, with percentage of successful placements in brackets, for Experiment A.
Waterlogged site |
Non-waterlogged site | ||||
Species |
Method |
Before rain |
After rain |
Before rain |
After rain |
Bluebush |
1. Mallen & Chaff, PVA, DDT and Rogor |
2.2aabcd (32)a |
0.9ab (17) |
2.3abc (22) |
2.labc (38) |
2. Mallen and Chaff |
3.lcd (34) |
2.6bcd (21) |
5.7de (42) |
6.1e (49) | |
3. Mallen without chaff |
0.2a (3) |
0.9ab (17) |
3.0bcd (40) |
12.5 (75) | |
4. Seed spread on cultivated ground |
|
|
|
| |
Wavy leaf saltbush |
1. |
1.0ab (37) |
2.3abcd (43 |
0.3ab (40) |
0.8ab (38) |
2. |
2.6bcd (45) |
1.6ab (41) |
1.4ab (44) |
0.4ab (26) | |
3. As above |
0.8ab (19) |
0.5ab (30) |
0.5 ab (20) |
0.4ab (29) | |
4. |
0 |
0 |
0 |
0 | |
River saltbush |
1. |
0a |
0.5ab (1) |
0a (0) |
0.6ab (3) |
2. |
0a |
0.3a (3) |
0.2ab (7) |
0.6ab (4) | |
3. As above |
0a |
4.4d (17) |
5.0cde (e) |
0a (0) | |
4. |
0 |
0 |
0 |
0 | |
aPlacement figures are for the January 1977 count, and establishment percentages are for the October 1976 count. For means in each site table the figures not next to the same letter are significantly different at P <0.5.
Table 2. Mean percentages of successful placements in Experiment B at two counting times, October 1976 (in brackets) and January 1977.
Species × seed |
Species × mulch | |||||
Treatment Rogor |
|
Species | ||||
| Species | Rogor | Nil | Mulch | M. brevifolia |
A. undulata |
A. rhagodioides |
M. brevifolia |
6.1 (9.4) aa |
46.7 (47.0) c |
Nil |
17.5 (34.2) ab |
62.5 (58.3) ef |
29.2 (20.0) bc |
A. undulata |
39.9 (33.0) c |
47.7 (31.7) c |
Chaff |
26.7 (28.3) bc |
12.5 (6.7) ab |
10.8 (6.3) a |
A. rhagofioides |
10.0 (4.5) a |
24.1 (16.4) b |
PVA |
10.8 (8.3) a |
50.0 (35.8) de |
14.2 (7.5) ab |
DDT |
39.2 (35.8) cd |
17.5 (12.5) ab |
14.2 (9.2) ab | |||
PVA and DDT |
37.7 (34.3) cd |
76.3 (48.3) f |
16.8 (9.3) ab | |||
a Results not beside the same letter in each table of means are significantly different < P 0.05.
The most outstanding result in Experiment A was the marked improvement in establishment achieved by using the Mallen niche seeder instead of spreading seed on cultivated land. The significance of the result is so clear-cut that the low results in the hand-spread plots were excluded from the statistical analyses. No significant effect of time of sowing was demonstrated on either site although there was a significant (P<0.01) interaction between species and sowing time.
The highest mean percentage establishment of viable seeds sown was 12.5 per cent for M. brevifolia Mallen niche-seeded without chaff on the non-waterlogged site. This result coincided with a mean of 75 per cent of successful placements.
In general the percentage of viable seeds giving plants proved to be a more sensitive index of establishment than the percentage of successful placements. Considering the former data chaff was found to give a significant (P<0.05) improvement in establishment of M. brevifolia and A. undulata in the pre-rain sowings, but not post-rain. This result applied only on the waterlogged site, and on the non-waterlogged site M. brevifolia gave the best establishment after rain without any mulch: There were marked differences in establishment by the different species with A. brevifolia giving the best establishment after rain without any mulch. A. rhagodioides gave generally poor results.
In Experiment B the coefficient of variation was much lower (45.1%) despite the fact that only percentage successful placement data were obtained. As shown in Table 2, Rogor reduced establihment significantly in M. brevifolia and A. rhagodioides and appeared to be detrimental to establishment of A. undulata. Establishment of M. brevifolia was significantly better with chaff treated with DDT or DDT and PVA, than without mulch if the assessment was made on the January 1977 establishment count. It is of considerable interest to note that the establishment counts in the unmulched plots of M. brevifolia were almost halved between the October 1976 and January 1977 counts. By contrast the mean establishment figures for all treatments for the two Atriplex spp. increased between the two counts, in some cases almost doubling. The mulch treatments enabled M. brevifolia to maintain its numbers and even to increase slightly between the two counting times. PVA treated chaff gave a significantly lower establishment than the other mulches with M. brevifolia, but with A. undulata only chaff with DDT and PVA was significantly better (P < 0.05). A. rhagodioides establishment was significantly better without mulch than with plain chaff. However, there was no significant difference in establishment between the unmulched result and that for any of the treated chaff treatments. Establishment of A. undulata averaged 76.3% of placements for the best treatment but M. brevifolia and A. rhagodioides reached only 39.2 and 29.2% respectively.
At the time of the first and second sowings soil samples were taken of the surface soil either in the niche or in the cultivated soil to gauge the salinity at the point where the seed was placed. The chloride contents of the samples were determined by titration and expressed as per NaCl. In the area of Experiment B the mean was 0.33 per cent NaCl for six samples taken in the niche in late May.
Adjacent to the experiment site, runs of the Mallen niche seeder were made in late April with chaff but no seed for soil sampling in late May. Samples of the top 10 mm of soil were taken from four sections of niche with and without chaff mulch. The former averaged 0.18% NaCl and the latter 0.70. The mean of eight niche samples in late April in the area of Experiment A was 0.37% NaCl and 0.24% in late May. In the conventional plots the corresponding figures were 1.46 and 0.79% NaCl.
The experiments presented in this paper are the first detailed field experiments conducted using the Mallen niche seeder. Because they represent only one season's results they must be regarded with some caution. However, the outstanding difference between niche seeded plots and conventionally sown plots is most encouraging. Further detailed experiments have been conducted more recently with the Mallen niche seeder and will be reported elsewhere. Results were sufficiently encouraging for several broadscale areas (about 10 ha each) to be sown in 1978. A. rhagodioides gave poor results but A. undulata gave 62.5% successful placements on the autumn sown site. The site on which this result was achieved was severely salt-affected and incapable of growing a profitable barley crop. The niche seeding technique has therefore been shown to be capable of producing a stand of salt-tolerant forage shrubs in a single operation.
The practicability of the niche seeding technique lies partly in the economic aspects of the method. The Atriplex spp. grow to a sufficient size for planting on a 3 x 3 metre grid to be acceptable. At such a spacing one bag of chaff is sufficient to treat one hectare of land. A closer spacing would be desirable with M. brevifolia if it were not for the fact that volunteer seedlings rapidly colonise the intervening areas. Denser stands could nevertheless be sown by using a closer spacing. The rates of seed used in the experiment for the treatments were approximately 0.4, 0.3 and 1.2 kg per ha for M. brevifolia, A. undulata and A. rhagodioides respectively.
The best number of seedlings as a percentage of viable seeds sown was 12.5% but most results were much lower. There is therefore considerable scope for improving establishment. The amount of seed used is not excessive and if cheap mechanical harvesting techniques can be developed, seed supply and cost will not present a problem for large-scale field sowing.
The results of the various treatment comparisons are confusing. There is no clear general benefit from the use of mulch or other treatments. There is scope for more detailed investigation of the effects of various mulches and time of sowing.
The limited account of soil sampling provided some extremely interesting results. In these experiments, on the soil type used, the niche seeding technique provided a much less saline environment for seed germination than normal cultivation. Moreover, the use of chaff resulted in a further substantial reduction in surface soil salinity between late April and late May. It is surprising in the light of these results that a clear response to chaff in terms of improved establishment was not obtained.
The Mallen niche seeder gave much better establishment of salt-tolerant forage shrubs on saline soil than was obtained by spreading seed on scarified ground. Time of sowing and various mulch treatments had significant effects on establishment. The effects were different on different species and there is a need for detailed work to determine the best planting technique for each species.
Malcolm, C.V. (1969). 'Use of halophytes for forage production on saline wastelands'. J. Aust. Inst. Agric. Science, 35: 32–49.
Malcolm, C.V. (1971). 'Establishing shrubs in saline environments West'. Aust., Dept. Agric. Tech. Bull. No. 14.
Malcolm, C.V. (1974). 'Forage production from shrubs on saline land'. J. Agric. West Aust., 15 (3) 4th Series 68–73.
Malcolm, C.V. (1979). 'Selection of shrubs for forage production from saline soils under natural rainfall'. International Conference on Indian Ocean Studies. Perth. W. Australia.
Malcolm, C.V.; Clarke, A.J. (1971). 'Collection and testing of forage plants for saline and arid areas'. Progress Report No. l., West. Aust. Dept. Agric. Tech. Bull. No. 8.
